Performance of capnometry in non-intubated infants in the pediatric intensive care unit

BACKGROUND: Assessing the ventilatory status of non-intubated infants in the Pediatric Intensive Care Unit (PICU) is a constant challenge. Methods to evaluate ventilation include arterial blood gas analysis (ABG), which is invasive and intermittent, and transcutaneous carbon dioxide monitoring (P(tcCO2)), which, while non-invasive, is also intermittent. A method that is non-invasive and continuous would be of great benefit in this population. We hypothesized that non-invasive capnometry via sidestream monitoring of exhaled carbon dioxide (CO(2)) would provide an acceptable measurement of ventilatory status when compared to ABG or P(tcCO2). METHODS: Preliminary prospective study of infants less than one year of age admitted to the PICU in a large urban teaching hospital. Infants not intubated and not requiring non-invasive ventilation were eligible. A sidestream CO(2) reading was obtained in a convenience sample of 39 patients. A simultaneous ABG was collected in those with an arterial catheter, and a P(tcCO2) was obtained in those without. RESULTS: Correlation of sidestream CO(2) with ABG was excellent (r(2) = 0.907). Sidestream correlated less well with P(tcCO2) (r(2) = 0.649). Results were not significantly altered when weight and respiratory rate were added as independent variables. Bland-Altman analysis revealed a bias of -2.7 with a precision of ±6.5 when comparing sidestream CO(2) to ABG, and a bias of -1.7 with a precision of ±9.9 when comparing sidestream CO(2) to P(tcCO2). CONCLUSIONS: Performance of sidestream monitoring of exhaled CO(2) is acceptable clinical trending to assess the effectiveness of ventilation in non-intubated infants in the PICU.